
ThetaTrap is a bounded Alpaca paper-options agent inspired by a YouTube idea: earnings can make premium expensive, so a defined-risk structure may benefit when volatility falls. Autonomy adds harder problems: stale quotes, inconsistent four-leg data, drifting model output, and ambiguous order timeouts. ThetaTrap therefore separates judgment from control. The frozen universe has seven eligible events and two explicit exclusions, DELL and NTAP. Featherless-hosted Qwen calls approved tools to inspect eligible opportunities and can veto risky event context. Deterministic Python builds one-contract iron condors, caps maximum loss at the lower of $500 or 0.5% of equity, and allows one broker entry attempt per strategy date after reviewing up to two candidates sequentially. Every Alpaca read and order uses the official MCP server. A persistent worker handles reconciliation, cancellation, and next-morning exit. ThetaTrap uses Alpaca Basic indicative data and simulated paper fills. It does not demonstrate proven profitability or live-capital readiness.
4 Sep 2026

Utilizing new frontier AI models comes at a cost... quite literally. Companies like Uber, Meta, and Microsoft have expended their AI resource budget due to internal over-use of AI.Thus, there is a growing need for enterprises to control AI spend without sacrificing user experience: not every task needs a premium proprietary model. This project aims to build that smart router by running variety of local models and making as few external Fireworks API calls as possible while still clearing the accuracy gate. Our agent will be tested across all eight categories: Factual knowledge, Mathematical reasoning, Sentiment classification, Text summarization, Named entity recognition, Code debugging, Logical reasoning, and finally Code generation.
13 Jul 2026